CN103415788B - Object detection device - Google Patents

Object detection device Download PDF

Info

Publication number
CN103415788B
CN103415788B CN201280011379.5A CN201280011379A CN103415788B CN 103415788 B CN103415788 B CN 103415788B CN 201280011379 A CN201280011379 A CN 201280011379A CN 103415788 B CN103415788 B CN 103415788B
Authority
CN
China
Prior art keywords
mentioned
detection
region
detection zone
image
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201280011379.5A
Other languages
Chinese (zh)
Other versions
CN103415788A (en
Inventor
野坂健郎
野坂健一郎
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Panasonic Intellectual Property Management Co Ltd
Original Assignee
Panasonic Intellectual Property Management Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Panasonic Intellectual Property Management Co Ltd filed Critical Panasonic Intellectual Property Management Co Ltd
Publication of CN103415788A publication Critical patent/CN103415788A/en
Application granted granted Critical
Publication of CN103415788B publication Critical patent/CN103415788B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01SRADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S3/00Direction-finders for determining the direction from which infrasonic, sonic, ultrasonic, or electromagnetic waves, or particle emission, not having a directional significance, are being received
    • G01S3/78Direction-finders for determining the direction from which infrasonic, sonic, ultrasonic, or electromagnetic waves, or particle emission, not having a directional significance, are being received using electromagnetic waves other than radio waves
    • G01S3/782Systems for determining direction or deviation from predetermined direction
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01SRADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S3/00Direction-finders for determining the direction from which infrasonic, sonic, ultrasonic, or electromagnetic waves, or particle emission, not having a directional significance, are being received
    • G01S3/78Direction-finders for determining the direction from which infrasonic, sonic, ultrasonic, or electromagnetic waves, or particle emission, not having a directional significance, are being received using electromagnetic waves other than radio waves
    • G01S3/782Systems for determining direction or deviation from predetermined direction
    • G01S3/783Systems for determining direction or deviation from predetermined direction using amplitude comparison of signals derived from static detectors or detector systems
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00Image analysis
    • G06T7/10Segmentation; Edge detection
    • G06T7/11Region-based segmentation
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00Image analysis
    • G06T7/10Segmentation; Edge detection
    • G06T7/174Segmentation; Edge detection involving the use of two or more images
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00Image analysis
    • G06T7/10Segmentation; Edge detection
    • G06T7/194Segmentation; Edge detection involving foreground-background segmentation
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/103Static body considered as a whole, e.g. static pedestrian or occupant recognition
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01VGEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS; TAGS
    • G01V8/00Prospecting or detecting by optical means
    • G01V8/10Detecting, e.g. by using light barriers
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01VGEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS; TAGS
    • G01V9/00Prospecting or detecting by methods not provided for in groups G01V1/00 - G01V8/00
    • G01V9/005Prospecting or detecting by methods not provided for in groups G01V1/00 - G01V8/00 by thermal methods, e.g. after generation of heat by chemical reactions
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/10Image acquisition modality
    • G06T2207/10048Infrared image
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/20Special algorithmic details
    • G06T2207/20212Image combination
    • G06T2207/20224Image subtraction
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V2201/00Indexing scheme relating to image or video recognition or understanding
    • G06V2201/07Target detection

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Electromagnetism (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Human Computer Interaction (AREA)
  • Multimedia (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • General Life Sciences & Earth Sciences (AREA)
  • Geophysics (AREA)
  • Geophysics And Detection Of Objects (AREA)
  • Radiation Pyrometers (AREA)
  • Image Processing (AREA)

Abstract

Provided is an object detection device which responds to changes in the ambient temperature and thus is unlikely to generate a false detection. An object detection device (1) is provided with: an acquisition unit (11) for acquiring a thermal image from a sensor device (2); a storage unit (12) for storing a background image; a detection unit (13) for detecting an object (3) to be detected; and an update unit (15) for updating the background image. The object detection device (1) detects an object (3) to be detected by means of the detection unit (13) and by means of a background difference method which detects the difference between the background image and an acquired image which is the present thermal image. For no-detection regions, in which an object (3) to be detected has not been detected, the update unit (15) updates the background image on the basis of the most recent acquired image. For detection regions, in which an object (3) to be detected has been detected, the update unit (15) finds an adjustment value from a temperature variation which has been updated by means of a first background update process, and uses the adjustment value to update the background image.

Description

Article detection device
Technical field
The present invention relates to a kind of using thermal image (acquisition image) obtain the difference of current thermal image and background image Lai The article detection device of detection object thing.
Background technology
In the past, there is provided a kind of thermal image from sensor device acquisition detection range simultaneously detects as right from thermal image The article detection device of the thermals source such as the people as thing.As by the method for article detection device detection object it is known that following background Calculus of finite differences:By pre-recorded for the thermal image that there is not the detection range in the moment of object in detection range for background image, Detect the difference of the current thermal image and background image being got by sensor device.Background subtraction have computing simple, Accuracy of detection also high such advantage under stable environment.
But, if the background image once obtaining in statu quo persistently is made using the article detection device of background subtraction With then the temperature in the detection range not relying on object (hereinafter referred to as " ambient temperature ") is possible to when changing produce Raw error detection is it is therefore desirable to be continuously updated background image.
Therefore, it is proposed to have the device (referring for example to patent documentation 1) of following function:Do not detecting from acquisition image In the case of going out object to be detected, update background image using this acquisition image.In addition it is also proposed that having following function Device (referring for example to patent documentation 2):Obtain separating the moment of stipulated time from the moment that have updated previous background image Take new image, the additive average value obtaining this acquisition image with current background image is to be updated to Background next time Picture.But, in the device described in patent documentation 2, in the case of detecting object from acquisition image, do not execute and add Method average treatment.
In addition, as from image identify human body device it is also proposed that there is the device of following function (referring for example to special Sharp document 3):Carry out the calculus of differences of background image and current acquisition image, and binaryzation is carried out with the threshold value of regulation, for Pixel below threshold value changes background image with current image.In device described in patent documentation 3, it is also equipped with following work( Energy:Even if pixel more than threshold value (that is, the pixel being detected by binaryzation) is present on current acquisition image, for It is judged as it not being the pixel of human body, also background image is changed with current acquisition image.
Patent documentation 1:Japanese Unexamined Patent Application 62-240823 publication
Patent documentation 2:Japanese Unexamined Patent Publication 10-308939 publication
Patent documentation 3:Japanese Unexamined Patent Publication 5-118916 publication
But, in the structure described in patent documentation 1,2, the image detect object does not carry out background image Renewal, therefore in detection range, the period of sustainable existence object does not carry out the renewal of background image, thus cannot tackle The situation that ambient temperature changes, it is possible to produce error detection.
Even if in addition, for not examining in the image detecting object device described in as patent documentation 3 Measure object (people) region carry out the renewal of background image in the case of, the region due to detecting object is not carried on the back The renewal of scape image, therefore also cannot response environment temperature change, it is possible to produce error detection.That is, for example in phase Although the body temperature of people is also correspondingly declined with ambient temperature when in the state of the lasting someone in same position, ambient temperature declines, but It is that the background image in region residing for people still keeps the high state of ambient temperature not to be updated.In this case, can sooner or later The body temperature of people is led to be less than threshold value with the difference value of background image, it is nobody's (leakage that article detection device is possible to error detection Report).On the contrary, when in the state of identical position continues someone, ambient temperature rises, the background image in the region residing for people Remain below the state of the ambient temperature of reality, that is, make one to have left this region, article detection device is also possible to mistake inspection Survey is in this region someone.
The present invention be in view of the above-mentioned origin of an incident and complete, its object is to provide a kind of by right with the change of ambient temperature Answer and do not allow to be also easy to produce the article detection device of error detection.
Content of the invention
The article detection device of the present invention is characterised by possessing:Acquisition unit, it obtains the temperature representing in detection range The acquisition information of distribution;Storage part, its storage represents above-mentioned detection range when there is not object in above-mentioned detection range The background information of interior Temperature Distribution;Test section, it, to be examined with respect to the change of above-mentioned background information according to above-mentioned acquisition information Survey and have or not object in above-mentioned detection range;And update section, it updates the above-mentioned back of the body being stored in above-mentioned storage part repeatedly Scape information, wherein, above-mentioned background information is divided into the region comprising to be detected object by above-mentioned test section by above-mentioned update section Detection zone and the no detection zone being made up of the region beyond this detection zone, for above-mentioned no detection zone, above-mentioned renewal Portion executes and is processed according to the first context update of the above-mentioned background information of above-mentioned acquisition information updating, for above-mentioned detection zone, on State update section execution to process using the second context update that corrected value updates above-mentioned background information, wherein, this corrected value is to be based on Temperature produced by above-mentioned first context update process is passed through in the extraction region that at least a portion of above-mentioned no detection zone is constituted Degree variable quantity is obtained.
In this article detection device it may be desirable to above-mentioned update section by above-mentioned no detection zone with above-mentioned detection zone Above-mentioned correction, as said extracted region, is obtained based on the temperature variation in this extraction region in the adjacent a part of region in domain Value.
In this article detection device it more desirable to be above-mentioned update section process institute according to by above-mentioned first context update The extent of deviation of the temperature variation of above-mentioned no detection zone producing switches between first mode and second mode, its In, in this first mode, using a part of region of the above-mentioned no detection zone of above-mentioned background information as said extracted region And above-mentioned corrected value is obtained based on the temperature variation in this extraction region, in this second mode, above-mentioned background information is upper The whole region stating no detection zone obtains above-mentioned school as said extracted region and based on the temperature variation in this extraction region On the occasion of.
The effect of invention
In the present invention, for detection zone, update section execution is using based at least a portion structure by no detection zone The corrected value that the temperature variation in the extraction region becoming is obtained is processed come the second context update to update background information, therefore passes through Corresponding with the change of ambient temperature and do not allow to be also easy to produce error detection.
Brief description
The purpose of the present invention and feature become clear and definite by the explanation of following accompanying drawing and preferred embodiment.
Fig. 1 is the synoptic diagram of the structure representing the article detection device involved by embodiment 1.
Fig. 2 is the flow chart of the action representing the article detection device involved by embodiment 1.
Fig. 3 is the explanatory diagram of the action representing the article detection device involved by embodiment 1.
Fig. 4 is the explanatory diagram of the action representing the article detection device involved by embodiment 1.
Fig. 5 is the explanatory diagram of the action representing the article detection device involved by embodiment 1.
Fig. 6 is the explanatory diagram of the action of the article detection device involved by other structures representing embodiment 1.
Fig. 7 is the explanatory diagram of the action of the article detection device involved by other structures representing embodiment 1.
Fig. 8 is the explanatory diagram of the action of the article detection device involved by other structures representing embodiment 1.
Fig. 9 is the explanatory diagram of the action representing the article detection device involved by embodiment 2.
Specific embodiment
Below, with reference to the accompanying drawing of the part forming this specification, embodiments of the present invention are illustrated in greater detail.Institute Have in accompanying drawing and be directed to same or similar partly additional same reference marker and omit the description.
As shown in figure 1, the article detection device 1 of present embodiment possesses:Acquisition unit 11, it obtains from sensor device 2 Information;Storage part 12, its storage information;Test section 13, it carries out the computing that background subtraction grades, and detects the presence of object 3;And Output section 14, it exports testing result.Here, people is set to object 3, but it is not limited to people, as long as becoming the object of thermal source Thing, the automobile for example travelling etc., the article detection device 1 of present embodiment just can be detected.
Sensor device 2 is with matrix by the sensor element such as such as thermopile element to infrared ray responsive (not shown) Shape carries out two-dimensional arrangement and is formed, and constantly to export table using two-dimentional thermal image (acquisition image) as temperature distribution image Show the information of the Temperature Distribution in detection range (sensing scope).Each sensor element is corresponding with each pixel of thermal image, below The Celsius temperature in the region detected by each sensor element is said as the pixel value of each pixel corresponding in thermal image Bright.Additionally, in the example in fig 1, article detection device 1 and sensor device 2 are mutually independent, but are not limited to this structure, also may be used To be provided integrally with sensor device 2 in article detection device 1.
The article detection device 1 of present embodiment is made up of computer, executes regulated procedure, cause by making computer The each function as acquisition unit 11, test section 13 and update section described later 15 realized by computer.Output section 14 includes monitoring dress Put, printing equipment etc. has the device of the function of testing result of assuming test section 13 to user.
Article detection device 1 passes through acquisition unit 11 and obtains, from sensor device 2, the Temperature Distribution representing in detection range Acquisition information (acquisition image), by test section 13 according to acquisition information with respect to the background information being stored in storage part 12 The change of (background image) has or not object 3 to detect in detection range.That is, article detection device 1 would indicate that in detection model The thermal image that there is not the Temperature Distribution of the detection range in moment of object 3 in enclosing is previously recorded in storage as background image In portion 12.This article detection device 1 termly obtains the Temperature Distribution of the detection range current as expression from sensor device 2 Thermal image acquisition image, obtain the background subtraction of the difference of image and background image by detection, come by test section 13 detection object things 3.
But, if the background image once obtaining in statu quo persistently is used, in the change due to the time period, refrigeration Heat etc. and when the temperature (hereinafter referred to as " ambient temperature ") that do not rely in the detection range of object 3 changes, object Detection means 1 is possible to produce error detection.
Therefore, the article detection device 1 of present embodiment possesses the background image that renewal is stored in storage part 12 repeatedly Update section 15.Here, update section 15 is being carried out examining based on the contrast of acquisition image and background image by test section 13 every time All update background image during the detection process surveying object 3, but be not limited to this structure, for example can also be by for often multiple The cycle that detection process are once updated updates background image.In addition, the moment reaching certain and predetermining for example can also be worked as When, update section 15 updates background image.
Below, the flow chart with reference to Fig. 2 updates the back of the body with the detection process of test section 13 detection object thing 3 and update section 15 The renewal of scape image illustrates the action of article detection device 1 centered on processing.In addition, Fig. 3~5 with the thermal image of 5 × 5 pixels are Example carrys out the process exemplified with article detection device 1.Additionally, below with the coordinate position (x, y) of each pixel in image with timely Carve the pixel value that t specifies each pixel in image.
Article detection device 1 first passes through acquisition unit 11 and obtains in the state of there is not object 3 from sensor device 2 The thermal image of output, each pixel value of this thermal image is prestored to storage as the initial value of background data Tb (x, y, t) In portion 12 (S1).Background data Tb (x, y, t) is the pixel value of each pixel constituting background image Tb.
Article detection device 1 obtains current temperature profile data T (x, y, t) by acquisition unit 11 from sensor device 2, And this temperature profile data T (x, y, t) is stored (S2) in storage part 12.Temperature profile data T (x, y, t) is to constitute to obtain The pixel value of each pixel of image T.
Test section 13 calculates, as Fig. 3 illustration, background data Tb (x, y, t) and the temperature being stored in storage part 12 The difference of distributed data T (x, y, t) is used as differential data Td (x, y, t) (S3).That is, representing differential data with formula 1 Td(x,y,t).Differential data Td (x, y, t) is the pixel value of each pixel constituting difference image Td.Test section 13 presses each picture Element is for the computing represented by all pixels progressive form 1 obtaining image T.
[formula 1]
Td (x, y, t)=T (x, y, t)-Td (x, y, t)
It is then detected that portion 13 carries out binaryzation (S4) using the threshold value of regulation to differential data Td (x, y, t).Now, examine Survey portion 13 carries out binaryzation to generate detection image I as Fig. 3 illustration as follows:By difference value for threshold value more than Pixel as " detection pixel " (S5) detecting object 3, difference value is less than the pixel of threshold value as not detecting object " no detecting pixel " (S6) of thing 3.In the example of Fig. 3~5, oblique line portion represent by detection pixel set constitute region, I.e. as detection zone D1 in the region detecting object 3, region representation in addition is by the set structure of no detection zone The region of one-tenth, i.e. no detection zone D0 as the region not detecting object 3.Constitute the picture of each pixel of detection image I Plain value I (x, y, t) is represented using threshold value C formula 2.
[formula 2]
That is, the pixel (x, y) of I (x, y, t)=1 is detection pixel, the pixel (x, y) of I (x, y, t)=0 is no to examine Survey pixel.Detection image I exports from output section 14 as the testing result of test section 13.C=2 in the present embodiment.
Then, article detection device 1 is carried out for the no detection being made up of the set no detecting pixel by update section 15 The first context update that region D0 updates background data Tb (x, y, t) processes (S7).Update section 15 is for no detection zone D0 All pixels, according to up-to-date temperature profile data T (x, y, t) correction be stored in background data Tb in storage part 12 (x, y, T), thus carry out the first context update process.Specifically, update section 15 calculates as Fig. 4 illustration and is stored in storage part Weighted average (the weighted of the background data Tb (x, y, t) in 12 and current temperature profile data T (x, y, t) Average) value, to carry out the first context update process.That is, no detection zone D0 new background data TbD0 (x, Y, t+1) represented using average weighted parameter alpha formula 3.In the example in fig. 4, setting parameter α=0.1.Update section 15 presses each picture Element is for the computing represented by all pixels progressive form 3 of no detection zone D0.
[formula 3]
TbD0 (x, y, t+1)=(1- α) TbD0 (X, y, t)+α TD0 (x, y, t)
Here, TbD0 (x, y, t) is the background data of no detection zone D0, TD0 (x, y, t) be no detection zone D0 work as Front temperature profile data.
Update section 15 is using the new background data TbD0 of no detection zone D0 being obtained by the first context update process (x, y, t+1), obtains corrected value used in the second context update described later is processed.Corrected value is based in background image Tb At least a portion by no detection zone D0 constitute extract region in by first context update process renewal after temperature Variable quantity obtain.In the present embodiment, using the whole region of no detection zone D0 as extraction region.Here, updating Portion 15 is obtained by all pixels that each pixel is directed to no detection zone D0 as (a) of Fig. 5 illustrates first and is represented with formula 4 Updated value Δ TbD0 (x, y, t) (S8).Updated value Δ TbD0 (x, y, t) is to represent in background image Tb to pass through the first background more The value of the variable quantity of pixel value in new no detection zone D0 processing after updating.
[formula 4]
ATbD0 (x, y, t) two TbD0 (x, y, t+1)-TbD0 (x, y, t)
Here, TbD0 (x, y, t) is the previous background data of no detection zone D0.
Then, update section 15 is obtained as the updated value Δ TbD0 in all pixels of no detection zone D0 extracting region The representative value of (x, y, t) is used as corrected value (S9).In the present embodiment, using as corrected value updated value Δ TbD0 (x, Y, t) representative value be set to the average value Δ Tav that represented with formula 5.But, representative value is not limited to meansigma methodss or central authorities Value, mode etc..Additionally, in formula 5, the prime number (pixel count) of wanting belonging to no detection zone D0 is expressed as | D0 |.
[formula 5]
In the example of fig. 5, in no detection zone D0 being made up of 19 pixels, updated value Δ TbD0 (x, y, t)= 0.1 pixel has 12, and therefore average value Δ Tav is 1.2/19 0.06.
Then, article detection device 1 is carried out for the detection zone being made up of the set of detection pixel by update section 15 The second context update that D1 updates background data TbD1 (x, y, t) processes (S10).Update section 15 is directed to all of detection zone D1 Pixel, is stored in detection zone D1 in storage part 12 using corrected value (average value Δ Tav) correction obtained by processing S9 Background data Tb (x, y, t) (below for TbD1 (x, y, t)), thus carry out the second context update process.Specifically, update Portion 15 will be stored in current background data TbD1 (x, y, t) and corrected value in storage part 12 as (b) of Fig. 5 illustrates It is added, to carry out the second context update process.That is, with formula 6 represent detection zone D1 new background data TbD1 (x, y,t+1).Computing as represented by all pixels progressive form 6 that each pixel is directed to detection zone D1 for the update section 15.
[formula 6]
TbD1(x,y,t+1)=TbD1(x,y,t)+ΔTav
In a word, update section 15 is directed to no detection zone D0 in background data Tb (x, y, t), the first background of progressive form 3 Renewal is processed, and the second context update for detection zone D1 progressive form 6 is processed.Update section 15 is using new by so obtain Background data Tb (x, y, t+1) change the background image Tb in storage part 12 at any time.As a result, storing pin in storage part 12 By value, the new background number of gained is updated to all pixels all pixels of detection zone D0 and this two side of detection zone D1 (no) It is used as background image Tb according to Tb (x, y, t+1).If additionally, there is not object 3 in the detection range of sensor device 2, Then difference image Td does not comprise detection zone D1, and whole region is no detection zone D0, therefore only passes through at the first context update Reason updates the whole region of background image Tb.
Afterwards, article detection device 1 returns to the process of S2, using update after background image Tb repeat S2~ The process of S10.
The article detection device 1 of present embodiment from the description above, even if be continuously present in same position in object 3 In the case that the period ambient temperature put there occurs change, update section 15 also can be for the detection zone detecting object 3 D1, follows the change of ambient temperature to update background image Tb.Thus, even if same long lasting for being present in object 3 In the case of position, also accordingly correctly update background image Tb with the change of ambient temperature, article detection device 1 passes through to make With the background image Tb after renewal, the detection process of the object 3 after can correctly carrying out.
In a word, update section 15, not only for no detection zone D0 not detecting object 3, is also directed to detect object 3 detection zone D1 updates background image Tb, and the whole region therefore, it is possible to make background image Tb corresponds to the change of ambient temperature Change.Even if for example in the state of identical position continues someone, ambient temperature declines, the background image Tb in the region of someone Also correspondingly it is updated with the change of ambient temperature, therefore article detection device 1 is it can be avoided that error detection is nobody's (leakage Report).In addition, when in the state of identical position continues someone, ambient temperature rises, article detection device 1 is it can be avoided that in people When leaving this region, error detection is in this region someone.
As a result, the article detection device 1 of present embodiment and the renewal not carrying out detection zone D1 in background image Tb Structure compare, have the advantages that by corresponding and do not allow to be also easy to produce error detection with the change of ambient temperature.
In addition, in the above-described embodiment, detection zone D1 is the pixel (detection pixel) that differential data is more than threshold value Set itself, but be not limited to this example it may be desirable to suitably expand detection zone D1.That is, detection zone D1 is only To be not limited to detect the region basis of object 3 including at least the region entirety being detected object 3 by test section 13 Body.Similarly, no detection zone D0 is the region beyond detection zone D1, is not limited to not detect the region of object 3 itself. That is, no detection zone D0 might be less that the region not detecting object 3.By expanding detection zone D1, can Overcome following problem.
That is, in poor situation of the pixel resolution of sensor device 2 etc., there is following situation:For in same pixel The pixel being mixed with the part that there is object 3 and there is not the part of object 3, the relatively low pixel value after output equalization (temperature), differential data is not for more than threshold value being included in no in detection zone D0.Physical presence object 3 in this pixel, Thus while pixel value is less than threshold value, but it is above original ambient temperature.Thus, according to the object detection of above-mentioned embodiment Device 1, background data is updated the temperature that portion 15 is updated to higher than original ambient temperature, thus becoming the former of error detection Cause.
On the other hand, in the case of expanding detection zone D1, in same pixel, it is mixed with the part that there is object 3 The pixel that there is not the part of object 3 is also contained in detection zone D1, therefore, it is possible to avoid the pixel by this pixel Value updates background data.Thus, it is possible to avoid background data to be updated portion 15 be updated to the temperature higher than original ambient temperature.
The expansion of detection zone D1 is processed and is specifically directed in detection image I being generated by above-mentioned process S4~S6 Pixel value I (x, y, t)=1 the expansion process (Dilation) that carries out of pixel.Thus, detection zone D1 of difference image Td It is not only the set of the pixel (detection pixel) that differential data Td (x, y, t) is more than threshold value as Fig. 6 illustrates, also expand Arrive greatly its periphery.Expansion process said here, as being common method for image processing techniquess, is by identical for value and phase The boundary line of the block of pixels connecting expands the process of a pixel laterally.By carrying out this expansion process, can will be same It is mixed with the part that there is object 3 in pixel and there is not the pixel of the part of object 3 and be included in detection zone D1.
Article detection device 1, will as Fig. 7 and Fig. 8 illustrates after the expansion carrying out detection zone D1 is processed Region beyond detection zone D1 as no detection zone D0 execution first context update process and second context update process (on State process S7~S10).Additionally, in the example of Fig. 6~8, oblique line portion represents detection zone D1, region representation in addition No detection zone D0.
(embodiment 2)
The article detection device 1 of present embodiment is obtaining the extraction using during the corrected value that the second context update is processed The setting in region is different from the article detection device 1 of embodiment 1.Below, for phase additional with embodiment 1 identical structure Same reference simultaneously suitably omits the description.
In the article detection device 1 of embodiment 1, using the whole region of no detection zone D0 as extracting region, more The representative value of updated value Δ TbD0 (x, y, t) of all pixels of no detection zone D0 is set at the second context update in new portion 15 The corrected value of reason.Therefore, the detection range in sensor device 2 is big and change of ambient temperature in detection range exists partially In the case of difference, it is possible to create following problem.Thermal source beyond object 3 as such as stove is present in away from object In the case of the position of thing (people) 3, the temperature meeting of sometimes unrelated with detection zone D1 original ambient temperature thermal source periphery Affect the corrected value that the second context update is processed.Thus, the background data of detection zone D1 is updated to than original sometimes The high temperature of ambient temperature.
On the other hand, in the present embodiment, update section 15 only will be adjacent with detection zone D1 in no detection zone D0 The a part of region connecing is as extraction region.That is, update section 15 calculate adjacent with detection zone D1 in no detection zone D0 Neighboring pixel in updated value Δ Tb (x, y, t) representative value (such as average value Δ Tav) as corrected value, using this correction Value carries out the second context update process.Specifically, update section 15 is directed to the expansion of detection image I execution n time (n=1,2 ...) Process, using the differential area of the region after expanding and detection zone D1 before expansion as extracting region, only will extract in region Pixel as object, obtain the representative value (corrected value) of updated value Δ Tb (x, y, t).Here, n is to determine to extract the wide of region The parameter of degree (width), as long as correspondingly carry out setting with the detection range of sensor device 2 and pixel resolution.
The article detection device 1 of present embodiment from the description above, due to only by no detection zone D0 with inspection Survey the adjacent a part of region of region D1 as extracting region, therefore, even if the change of ambient temperature in detection range is deposited It is also possible to calculate the corrected value that suitable value is processed as the second context update in the case of deviation.That is, passing through The second context update using corrected value is processed, and is updated using the change of the ambient temperature near object 3 and detects object The background data of detection zone D1 of thing 3, therefore this background data are corrected as more accurately being worth.Thus, even if in detection model Enclose interior ambient temperature change there is deviation in the case of it is also possible to avoid detection zone D1 in the second context update is processed Background data be updated to the temperature higher than original ambient temperature.This structure is big for the detection range of sensor device 2 The situation that the change of the ambient temperature in detection range such as situation easily produces deviation is particularly useful.
In addition, in the case of the block that there are multiple interconnective detection pixels as Fig. 9 illustration, update section 15 Detection image Im (m=1,2 ...) only existing each piece can also be generated by each block is interim, for each detection image Im individually Set and extract region.That is, update section 15 generates corresponding detection image Im of quantity with the block of detection pixel, for each Detection image Im is executed separately n (being twice in Fig. 9) expansion process, by the region after expanding and the detection zone before expansion The differential area of D11, D12 is as extraction region D13, D14.Additionally, in fig .9, with the direction of oblique line to as detection pixel Detection zone D11, D12 of block and as corrected value calculating object extraction region D13, D14 be distinguish between.
In this case, update section 15 is directed to and extracts region D13, D14 calculating corrected value, is updated each using each corrected value The background data of detection zone D11, D12.I.e., in the example of figure 9, calculated based on the extraction region D13 of detection image I1 Corrected value be used for detection image I1 detection zone D11 background data update, the extraction region D14 based on detection image I2 The background data that the corrected value calculating is used for detection zone D12 of detection image I2 updates.
Additionally, in the present embodiment, as the determining method extracting region, simply employ in image processing techniquess Expansion process, but the configuration according to sensor device 2, be possible in real space with the detection adjacent pixel of pixel Not near object 3.It is the situation of such configuration, for example sensor filled in the configuration being known a priori by sensor device 2 Put 2 configure obliquely downward in the case of, update section 15 can also be only using the connecing as object 3 and ground in detection pixel The neighboring pixel in the foot region of contact is as extraction region.
Other structures and function are identical with embodiment 1.
(embodiment 3)
The article detection device 1 of present embodiment is obtaining the extraction using during the corrected value that the second context update is processed The setting in region is different from the article detection device 1 of embodiment 2.Below, for phase additional with embodiment 2 identical structure Same reference simultaneously suitably omits the description.
In the present embodiment, update section 15 can be only will be adjacent with detection zone D1 in no detection zone D0 A part of region as extract region first mode and using the whole region of no detection zone D0 as extract region second Switch between this two action patterns of pattern.Update section 15 is no detected according to produced by processed by the first context update The extent of deviation of the temperature variation of region D0, switches between these first mode and second modes.
Specifically, update section 15 judges updated value Δ TbD0 (x, y, t) calculating with the first context update process Value be directed to no detection zone D0 all pixels whether generally uniform, if it is determined that uneven, then enter in the flrst mode Action is made, if it is determined that uniformly, then carries out action under the second mode.Here, update section 15 for example will no detection zone D0 In the difference value of maxima and minima of updated value Δ TbD0 (x, y, t) and regulation threshold value be compared, if difference Value is less than threshold value, then generally uniform, more than threshold value, is then judged as uneven.It is not limited to this structure, update section 15 is for example The standard deviation of updated value Δ TbD0 (x, y, t) can also be obtained, whether be in the range of the threshold value of regulation according to this value, come Judge whether generally uniform.
The article detection device 1 of present embodiment from the description above, because the temperature according to no detection zone D0 becomes The extent of deviation of change amount changes the extraction region of the calculating object as corrected value, and therefore update section 15 can be by optimal side Method calculates corrected value.
That is, using a part of region adjacent with detection zone D1 as extract region first mode in detection range The change of ambient temperature is effective in the case of there is deviation, it is possible that producing because as the pixel count calculating object Reduce and cause the deviation of corrected value, along with expansion process computational load increase.On the other hand, will no detection zone Although the change that the whole region of D0 is unsuitable for ambient temperature in detection range as the second mode extracting region exists partially The situation of difference, but there is deviation, the companion that will not produce the corrected value causing because of the minimizing as the pixel count calculating object Advantage with the increase of the computational load of expansion process.The update section 15 of present embodiment is according to the temperature in no detection zone D0 The extent of deviation of degree variable quantity switches between these first mode and second modes, therefore can pass through optimal all the time Method calculates corrected value.
Additionally, update section 14 be set in the first mode to extract region region be not limited in no detection zone D0 with inspection Survey the adjacent a part of region of region D1, as long as a part of region of no detection zone D0.
Other structures are identical with embodiment 2 with function.
It is explained above the preferred embodiment of the present invention, but the invention is not restricted to these specific embodiments, energy Enough carry out the various changes without departing from claim categories and deformation, it falls within scope of the invention.

Claims (3)

1. a kind of article detection device is it is characterised in that possess:
Acquisition unit, it obtains the acquisition information representing the Temperature Distribution in detection range;
Test section, it with respect to the change of background information, to detect and to have or not in above-mentioned detection range according to above-mentioned acquisition information Object, this background information represents that the temperature in the above-mentioned detection range when there is not object in above-mentioned detection range is divided Cloth, and be stored in storage part;And
Update section, it updates the above-mentioned background information being stored in above-mentioned storage part repeatedly,
Wherein, above-mentioned background information is divided into the detection comprising the region being detected object by above-mentioned test section by above-mentioned update section Region and the no detection zone being made up of the region beyond this detection zone,
For above-mentioned no detection zone, above-mentioned update section execution is according to first back of the body of the above-mentioned background information of above-mentioned acquisition information updating Scape renewal is processed, and for above-mentioned detection zone, above-mentioned update section execution updates second back of the body of above-mentioned background information using corrected value Scape renewal is processed, and wherein, this corrected value is to pass through based on by the extraction region that at least a portion of above-mentioned no detection zone is constituted Temperature variation produced by above-mentioned first context update process is obtained.
2. article detection device according to claim 1 it is characterised in that
The a part of region adjacent with above-mentioned detection zone in above-mentioned no detection zone is carried by above-mentioned update section as above-mentioned Take region, above-mentioned corrected value is obtained based on the temperature variation in this extraction region.
3. article detection device according to claim 1 and 2 it is characterised in that
The temperature variation of above-mentioned update section above-mentioned no detection zone according to produced by processed by above-mentioned first context update Extent of deviation switch between first mode and second mode, wherein, in this first mode, by above-mentioned background information A part of region of above-mentioned no detection zone obtain as said extracted region and based on the temperature variation in this extraction region Above-mentioned corrected value, in this second mode, the whole region of the above-mentioned no detection zone of above-mentioned background information is carried as above-mentioned Take region and above-mentioned corrected value is obtained based on the temperature variation in this extraction region.
CN201280011379.5A 2011-03-17 2012-03-14 Object detection device Active CN103415788B (en)

Applications Claiming Priority (3)

Application Number Priority Date Filing Date Title
JP2011-059699 2011-03-17
JP2011059699A JP5799219B2 (en) 2011-03-17 2011-03-17 Object detection device
PCT/IB2012/000476 WO2012123808A1 (en) 2011-03-17 2012-03-14 Object detection device

Publications (2)

Publication Number Publication Date
CN103415788A CN103415788A (en) 2013-11-27
CN103415788B true CN103415788B (en) 2017-02-15

Family

ID=46830097

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201280011379.5A Active CN103415788B (en) 2011-03-17 2012-03-14 Object detection device

Country Status (6)

Country Link
US (1) US9189685B2 (en)
EP (1) EP2672294A1 (en)
JP (1) JP5799219B2 (en)
KR (1) KR20140007402A (en)
CN (1) CN103415788B (en)
WO (1) WO2012123808A1 (en)

Families Citing this family (17)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103575323B (en) * 2012-07-30 2016-03-30 日电(中国)有限公司 Take detection method and device
CN109974244B (en) * 2014-02-17 2021-01-26 松下电器产业株式会社 Control method for air conditioner
JP6376804B2 (en) * 2014-04-01 2018-08-22 キヤノン株式会社 Image forming apparatus, image forming apparatus control method, and program
JP2015216482A (en) * 2014-05-09 2015-12-03 キヤノン株式会社 Imaging control method and imaging apparatus
JP6415178B2 (en) 2014-08-19 2018-10-31 キヤノン株式会社 Printing apparatus and data updating method
JP6374757B2 (en) * 2014-10-21 2018-08-15 アズビル株式会社 Human detection system and method
TWI554747B (en) * 2014-12-04 2016-10-21 台達電子工業股份有限公司 Human detection system and human detection method
CN105423494B (en) * 2015-12-11 2018-05-01 四川长虹电器股份有限公司 A kind of bearing calibration and air-conditioning equipment
CN107233082B (en) * 2016-03-29 2020-04-21 广州斯摩莱信息科技有限公司 Infrared thermal imaging detection system
JP6654091B2 (en) * 2016-04-19 2020-02-26 アズビル株式会社 Monitoring device, monitoring method, and program
JP7228767B2 (en) * 2018-05-16 2023-02-27 パナソニックIpマネジメント株式会社 SENSOR SYSTEM, AIR CONDITIONING SYSTEM, OBJECT DETECTION METHOD AND PROGRAM
KR20190138325A (en) * 2018-06-05 2019-12-13 휴렛-팩커드 디벨롭먼트 컴퍼니, 엘.피. Image forming apparatus to detect user and method for controlling thereof
US11162015B2 (en) 2020-02-14 2021-11-02 Halliburton Energy Services, Inc. Geopolymer formulations for mitigating losses
US11332654B2 (en) 2020-02-14 2022-05-17 Halliburton Energy Services, Inc. Well bore spacer and efficiency fluids comprising geopolymers
US11242479B2 (en) * 2020-02-14 2022-02-08 Halliburton Energy Services, Inc. Geopolymer cement for use in subterranean operations
JP7418652B2 (en) * 2021-02-19 2024-01-19 三菱電機株式会社 Human detection device, electrical equipment, human detection method, and human detection system
CN113743222A (en) * 2021-08-04 2021-12-03 北京澎思科技有限公司 Body temperature measuring method and device, electronic equipment and readable storage medium

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0933662A (en) * 1995-07-21 1997-02-07 Murata Mfg Co Ltd Detecting equipment of object generating infrared rays
JP2002148354A (en) * 2000-11-07 2002-05-22 Matsushita Electric Ind Co Ltd Human body detector
CN101169891A (en) * 2006-11-30 2008-04-30 中国科学院长春光学精密机械与物理研究所 Intelligent anti-theft monitor
US7693331B2 (en) * 2006-08-30 2010-04-06 Mitsubishi Electric Research Laboratories, Inc. Object segmentation using visible and infrared images
CN101702035A (en) * 2009-02-19 2010-05-05 黄程云 Digital quasi-static passive human body detector

Family Cites Families (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS62240823A (en) 1986-04-14 1987-10-21 Matsushita Electric Works Ltd Image sensor for monitor control
JPH05118916A (en) 1991-10-25 1993-05-14 Matsushita Electric Ind Co Ltd Human body recognition apparatus
JP3453870B2 (en) * 1994-09-21 2003-10-06 松下電器産業株式会社 Image processing device and applied equipment using the image processing device
DE69616191T2 (en) 1995-07-19 2002-03-14 Matsushita Electric Industrial Co., Ltd. Movement pattern recognition device for determining the movement of people and for counting people passing by
JPH0962822A (en) * 1995-08-28 1997-03-07 Matsushita Electric Ind Co Ltd Human body movement detection device and detection device for number of passing peoples
JPH10308939A (en) 1997-05-08 1998-11-17 Nec Corp Infrared monitoring system
US7415164B2 (en) * 2005-01-05 2008-08-19 Mitsubishi Electric Research Laboratories, Inc. Modeling scenes in videos using spectral similarity
JP5103767B2 (en) * 2006-03-27 2012-12-19 日産自動車株式会社 Temperature detection device
US8300890B1 (en) * 2007-01-29 2012-10-30 Intellivision Technologies Corporation Person/object image and screening
JP4894002B2 (en) * 2007-03-12 2012-03-07 サクサ株式会社 Moving body detection device and moving body detection system
US8446468B1 (en) * 2007-06-19 2013-05-21 University Of Southern California Moving object detection using a mobile infrared camera
US8934020B2 (en) * 2011-12-22 2015-01-13 Pelco, Inc. Integrated video quantization
JP5948983B2 (en) * 2012-03-09 2016-07-06 オムロン株式会社 Image processing apparatus, image processing method, and image processing program

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0933662A (en) * 1995-07-21 1997-02-07 Murata Mfg Co Ltd Detecting equipment of object generating infrared rays
JP2002148354A (en) * 2000-11-07 2002-05-22 Matsushita Electric Ind Co Ltd Human body detector
US7693331B2 (en) * 2006-08-30 2010-04-06 Mitsubishi Electric Research Laboratories, Inc. Object segmentation using visible and infrared images
CN101169891A (en) * 2006-11-30 2008-04-30 中国科学院长春光学精密机械与物理研究所 Intelligent anti-theft monitor
CN101702035A (en) * 2009-02-19 2010-05-05 黄程云 Digital quasi-static passive human body detector

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
智能车辆系统车辆流量检测技术的研究;代建辉;《中国优秀硕士学位论文全文数据库 工程科技II辑》;20090515(第05期);第三章 *
运动车辆的检测识别与跟踪;朱明旱;《中国优秀硕士学位论文全文数据库 工程科技II辑》;20060515(第05期);第二章 *

Also Published As

Publication number Publication date
WO2012123808A1 (en) 2012-09-20
US9189685B2 (en) 2015-11-17
JP5799219B2 (en) 2015-10-21
CN103415788A (en) 2013-11-27
EP2672294A1 (en) 2013-12-11
US20130329959A1 (en) 2013-12-12
JP2012194121A (en) 2012-10-11
KR20140007402A (en) 2014-01-17

Similar Documents

Publication Publication Date Title
CN103415788B (en) Object detection device
Yuan et al. Accurate displacement measurement via a self-adaptive digital image correlation method based on a weighted ZNSSD criterion
Lin et al. Automated body feature extraction from 2D images
US9928411B2 (en) Image processing apparatus, image processing system, image processing method, and computer program product
CN112287759A (en) Tumble detection method based on key points
CN100544622C (en) Data processing method for robot tactile sensing information syncretizing
CN111179279A (en) Comprehensive flame detection method based on ultraviolet and binocular vision
CN103345632B (en) A kind of battery tail end surface scratch defect inspection method
CN106203370B (en) A kind of test near and distance system based on computer vision technique
CN106153098A (en) The wearing state detection method of a kind of wearable device and device
CN108335454B (en) A kind of fire behavior detection method and device
CN111189494A (en) Measurement label, image color restoration method thereof and measurement identification method
JP6262105B2 (en) Image processing apparatus, image processing system, image processing method, and program
Bei et al. Sitting posture detection using adaptively fused 3D features
CN104603838A (en) Apparatus and method for monitoring object from captured image
CN109632825A (en) A kind of automatic testing method of coil of strip surface abnormalities protrusion
CN105957087A (en) Spray shape detection method and device
MX2020002349A (en) Automated control and inspection system for manufacture and measurement of apparel.
CN103150558A (en) Machine vision-based display terminal operation response matching detection method
BR112019016903A2 (en) methods for measuring the shape and for building a three-dimensional model of the inner surface of hollow objects and device for measuring the shape
CN108961384B (en) Three-dimensional image reconstruction method
KR101426864B1 (en) Region-based edge enhancement method for short-distance thermal target tracking
CN106123808B (en) A method of it is measured for the deflection of automobile rearview mirror specular angle degree
CN109102486A (en) Detection method of surface flaw and device based on machine learning
JP2015228992A (en) Visual axis analysis system and visual axis analysis device

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C41 Transfer of patent application or patent right or utility model
TA01 Transfer of patent application right

Effective date of registration: 20150911

Address after: Osaka Japan

Applicant after: PANASONIC INTELLECTUAL PROPERTY MANAGEMENT Co.,Ltd.

Address before: Osaka Japan

Applicant before: Matsushita Electric Industrial Co.,Ltd.

C14 Grant of patent or utility model
GR01 Patent grant